Slashdot Mirror
Building a Large Linux Knowledgebase
linuxfan writes "It looks like LinuxQuestions.org is aiming to build the largest independent Linux-related knowledgebase using a Wiki. They are using the same software as Wikipedia (MediaWiki), are using a Creative Commons license and look to be off to a good start."
The one thing keeping me from switching to Linux has a topic, but no one has posted any information there.
Does linux gaming work, or do you just have to play Neverwinter Nights over and over? (reply some good stuff here, and stick it in the wiki, too).
"If you think you have things under control, you're not going fast enough." --Mario Andretti
The only problem with the Linux Documentation Project is that it tends to have a lot of outdated information there.
The Wikipedia should be more up-to-date.
Then again - I just Google. . . . .
This sounds like a good idea. However, there are two issues that would hinder something like this:
1. Variability between Linuxes. There are many distros out there and they all have their own ways of doing the same things.
2. Variability within Linuxes. Different distros also change their commands between versions. Any knowledge-base specific enough to be of help would have to be extremely thorough. A person working with a 2.4.22-gentoo-r3 kernel, for example, might have a different kernel than someone with the vanilla 2.4.22.
However, if enough knowledgeable people use the system, then it will hopefully conquer these problems and be useful. But for now, see the gentoo forums for what open source documentation done by a community is capable of. (However, this is only within one distro that is relatively new so it has an easier time of things.)
2.
A NYC lawyer blogs. http://www.chuangblog.com/
Without wishing to sound too sarcastic, you are going to need a tad more than 19 articles to convince me you have a valid knowledge repository. I'll stick with the Linux Documentation Project for the moment thanks
Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
What truth?
There is no dupe
Wiki's seem to bring together the reasons why the web is seen as useful (disregarding the free pr0n). Easy content creation, accessibility and ease of use. With wiki's you get the added benefit of a central repository to look for the information you need.
As wiki's grow and become known the need for search engines might lessen. The first resort for information will change from google to the relevan wiki. Google will be seen as second in importance. After all, wiki has the structure of yahoo with the benefits of quality content.
I know this a stab at being funny, but...
The truth is pretty simple. Google's not enough. Sometimes for the most obscure questions, either most people consider you a nub and tell you to RFTM (which, in many cases, no manual exists), or they don't know themselves how to solve the problem. So, we often spend hours and hours going through google to find out how to get X hardware to work on Y configuration to find that Alan Cox had brilliantly broken my soundcard in 2.4.20... These are the kinds of things that there really should be a one-stop place to find what you need..
Surprisingly, it's taken this long for anyone to realize this...
"Victory means exit strategy, and it's important for the President to explain to us what the exit strategy is." G.W.Bush
Comment removed based on user account deletion
The difference is that this is a wiki. Anyone can update the wiki. Updating LDP docs ain't that easy - and goes through a lot of (and much needed) review cycles which takes quite some time before it is actually published.
This wiki knowledgebase idea is novel, in that it will allow more docs, pushed to the users in much more rapid fashion.
Yes, but that's what they* want you to know and when they* want you to know it.
*'They' being those who buy words on Google.
A feeling of having made the same mistake before: Deja Foobar
For example, there currently is only _one_ Chinese support on Linux FAQ, and it not only does not talk about Simplified Chinese, which is, last time I checked, a standard, but is terribly, terribly outdated.
The very structure of LinuxQuestions.org may allow them to produce documentation that is much more recent. While tLDP is all contributed articles by anyone who had some time to write something, LinuxQuestions is on the basis of Q&A. If they make a knowledgebase out of just questions and the most helpful response, it could very well be more useful in terms of bredth than tLDP, though probably not as much so in depth.
A wiki would solve this problem, since it is so easy to add information.
I've read many howtos with information that has been slightly wrong, a path has changed, a new configuration option has been added. Usually it has been quite obvious to see how to adapt the howto to the current situation. I've never reported any of these slight mistakes to the maintainers (yeah, lazy me...) but with a wiki the threshold would be much lover.
I'll be sure to help with what little I know.
Does it strike anyone as a bit of a negative attribute that you have to have a massive knowledgebase to use your operating system?
Microsoft has a knowledgebase as well, but only for troubleshooting, service packs, and development. I've been to their site maybe twice.
However, I've always had to rely on poorly-written HOWTOs and other documents to spend three hours just getting a sound card or USB mouse to work under X.
Call me crazy, but needing a huge database of tutorials to actually get things up and running is not exactly something you should be shouting from the rooftops.
"Sufferin' succotash."
Say you need to make a file over 2GB (very common in video processing) and you stumble across this page. You come away thinking linux doesn't support large files, when really it does. In this case you can scroll all the way to the bottom and the author is nice enough to have timestamped the page - 1998. But still, how do you know whether the old information is outdated? Even if you choose some arbitrary cutoff date ("information after Jan 1, 2003 is likely to be fresh") how do you tell google to only find information after that date? You can't.
Unfortunately there's no easy solution to the problem of outdated information. Useful documentation takes attention and manpower. But perhaps an easily modifiable Wikki-format will encourage more updates and more participation than comparable efforts such as The Linux Documentation Project, which is really just a smattering of FAQs, HOWTO-s, guides, and man pages with no real coherence - full of duplication and stale information.
This sort of theme is becoming increasingly prevalent these days. Take a look at ESR's recent article on trying to get CUPS to work. The bottom line is, rather than wasting energy writing more HOWTO's to get some impossibly-difficult-to-install software working, how about spending the time making the software easier to install?? Hmmmm? Or is Linux doomed to be used only by elitists?
How is linux ever going to gain any ground on the desktop if linux advocates are just going to bury their head in the sand and pretend that microsoft = bad/evil, and linux = perfection. He makes a fucking legit point.
I hope this post comes up for me to meta-mod.
Mind explaining why my post is "Flamebait?"
I simply made a point; an opinion of my own. If you disagree with it, reply or disregard, but don't mod it down. What makes Linux so great that it's above criticism?
"Sufferin' succotash."
One of the most frustrating things about Linux distributions is conflicting documentation and obselete and current documentation mixed into the same directories. I'll look at some documentation for packet filtering, for example, and be left totally bewildered as to which methods and software are actually current and intended for use and which ones are considered outmoded (and, afterwards, I run off and just use OpenBSD).
It would be a great help if some of the distribution maintainers contributed to the wiki. They can say things like, "Debian 3.0 uses software X for doing Y, while Debian 4.0 uses Z, a replacement for X." Some one else can say, "Well, Fedora chose to use W for doing Y, so you have to do this this and this differently."
Eliminating ambiguity can be, perhaps, the benefit of a "real-time" wiki.
Vote in November. You won't regret it.
Wikipedia gets enough traffic, and has enough regular users, that all changes are reviewed by at least a couple people, in practice; that's all that's needed to keep vandalism manageable, it appears. On a few obscure pages I've authored, mass deletions and nasty comments were removed within hours; the higher traffic pages are even more monitored, to the extent that 'editing wars' break out over them, requiring editors to lock pages for a period of time to let tempers cool.
So what the linux questions wiki needs for reliability is a critical mass of users; the rest takes care of itself.
Anyone who loves or hates any language, platform, or manufacturer, doesn't know what they're talking about.
Hmm. The Germany government has been supporting OpenFacts for some time. I guess I've tried to get people interested in that site, but it hasn't gotten very far (in the English version, at least -- the Germans seem to be doing well in the Deutsch version)
The good news for the new project is that all OpenFacts material is public domain, so it's fair to cut-n-paste off that site (well, if there's anything very useful).
Um, and how is this site run by LinuxQuestions independent, exactly? I guess I don't quite understand how that can be the case. Whatever, I guess.